Probabilistically checkable proofs

Author(s)
Sudan, Madhu
Thumbnail
DownloadSudan_Probabilistically checkable.PDF (203.2Kb)
PUBLISHER_POLICY
Terms of use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Metadata
Show full item record
Abstract
Can a proof be checked without reading it? That certainly seems impossible, no matter how much reviewers of mathematical papers may wish for this. But theoretical computer science has shown that we can get very close to this objective! Namely random consistency checks could reveal errors in proofs, provided one is careful in choosing the format in which proofs should be written. In this article we explain this notion, constructions of such probabilistically checkable proofs, and why this is important to all of combinatorial optimization.
Date issued
2009-03
URI
http://hdl.handle.net/1721.1/52308
Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory; Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Journal
Communications of the ACM
Publisher
Association for Computing Machinery
Citation
Sudan, Madhu. “Probabilistically checkable proofs.” Commun. ACM 52.3 (2009): 76-84.
Version: Author's final manuscript
ISSN
0001-0782
Collections